determinant india

8/7/2019 Determinant india

1/41

NCAERWorking Paper Series No. 85

Determinants of Maternal Health

Care Utilisation in India:

Evidence from a Recent Household Survey

Abusaleh Shariff

Geeta Singh


2/41

NCAER

NCAER Working Papers primarily disseminate the results of research work

done at or for the National Council of Applied Economic Research. Theycan be cited or quoted only with full acknowledgement of the Council. Theviews expressed are those of the authors and do not imply endorsement

by the Council.

Abusaleh Shariffiswith the National Council of Applied Economic Research,New Delhi 110 002.Geeta Singh is with the Harvard Institute of International Development, HarvardUniversity, Cambridge, USA.

This paper was prepared for the Programme of Research on Human

Development of the National Council of Applied Economic Research

sponsored by the United Nations Development Programme


3/41

NCAER

National Council of Applied Economic Research, February 2002

All rights reserved. No part of this publication may be reproduced,

stored in a retrieval system or transmitted in any form or byany means, electronic, mechanical, photocopying,


4/41

NCAER

ABSTRACT

In India, utilisation of basic health services has remained poor. The reasons may be low levels of

household income, high illiteracy and ignorance, and a host of traditional factors. In this paper an

attempt is made to discuss the issues associated with the demand and supply of the five measures of

maternity care-antenatal care, blood pressure check up, place of delivery, use of trained help at the

time of delivery and postnatal care. Econometric analysis is undertaken to find out the determinants of

the use of reproductive health care services among rural Indian households. Rural data from 32,000

households belonging to 1765 villages across all parts of India was collected by National Council ofApplied Economic Research in 1994. The multi-model survey was conducted both at the national and

state level. The analysis pertains to 7635 women in the reproductive age group who delivered a child

in the year before the survey. The focus on the role of education, information and economic factors as

determinants of health care accessibility and their utilisation is the speciality of this analysis. Analysis

shows that education and information variables significantly increase the utilisation rates for prenatal,

child delivery and postnatal health care. Women with primary education are more likely to use

maternal health services as compared to illiterate women, even after controlling for income and health

care supply factors. Exposure to media increases the probability of reproductive health care

utilisation. Economic factors such as wages and income are important only for the utilisation of child

delivery services. Access to locally available health services significantly increases maternity care

use. An important health care facility in this respect has been the village level ICDS centre. Further,

probit regressions analysis is used to examine the impact of individual, household and community

level variables on the above choices of reproductive health care.


5/41

NCAER

CONTENTS

1. Introduction 12. Reproductive Health Services 4

3.Data and Methodology 11

4.Empirical Analysis 15

5. Conclusions 24

References 26Tables 27


6/41

NCAER1. INTRODUCTION

In most developing countries, such as India, utilisation of basic health services has remained

poor even though there has been increasing public and private expenditure on the provision of

advanced health care. The low utilisation seems to be due to low levels of household income,

high illiteracy and ignorance, and a host of traditional factors. On the other hand, despite

substantial public investments in health infrastructure the supply of such services continues to

be inadequate and of poor quality. In addition, several inefficiencies such as an over emphasis

on secondary and tertiary hospital care, skewed distribution of health services favouringurban areas (Nayler et al., 1999) and gender discrimination in access to health care are all

pervasive (Shariff, 1999). Recent years have seen a rapid expansion of those health care

services that have large private but low social returns. There has also been an excessive

reliance on physicians rather than paramedics in the health care delivery mechanisms.

These weaknesses of the health care system have also affected the provision of

reproductive health especially of the natal-mother and childcare. In addition, the public

policy in this area has focussed on the use of contraceptives and has failed to address other

reproductive health needs of women. The existing literature on the use of reproductive health

services has focused on the effectiveness of various family planning methods, while studies

pertaining to the accessibility and use of services for the management of pregnancy and

childbirth are rare. As is well known, the health of the foetus or the newborn child affects its


7/41

NCAERmortality associated with childbirth and high infant mortality are observed in rural India

along with low rates of antenatal and postnatal care utilisation. In addition, the use of these

services varies considerably across socio-economic groups and geographical regions. These

variations exist partly due to differences in health seeking behaviour across the groups,

determined by economic, social and cultural factors and partly due to differences in access to

health care facilities.

In the developing economies, especially those that are large and diverse, many studies

have pointed out important benefits accruing from providing basic education and information

through mass media in terms of improvement in a number of human development parameters.

This paper focuses on the role of education, information and economic factors in the use of

selected reproductive health care services as well as the effect of health care accessibility on

their utilisation. Research on health outcomes has shown the significant positive effect of

information and education on health care use and health outcomes. The following analysis

shows that education and information variables significantly increase the utilisation rates of

prenatal, child delivery and postnatal health care. Women with primary education are, on an

average, 10 per cent more likely to use maternal health services relative to illiterate women,

even after controlling for income and health care supply factors. Exposure to media, such as,

radio, television or the printed media, increases the probability of reproductive health careutilisation by about 4 per cent.


8/41

NCAERby a description of the data and basic statistics in Section 3. Section 4 presents and discusses

the results from the empirical exercise. The last section concludes the paper.


9/41

NCAER2. REPRODUCTIVE HEALTH SERVICES

There exists a large network of institutions in the country for the delivery of health and

family welfare services at the community and higher levels set up by the state and national

governments. Although the emphasis has been on the family planning aspect of the

reproductive health services, the existing health care network provides other kinds of

maternity care services that are analysed in this paper. The majority of these services are

provided at the community level through various types of health workers (Figure 1). As a part

of the outreach program of delivery of health care services, there are female multi-purpose

workers stationed at health sub-centres that cater to a group of contiguous villages. However,

Dais (midwives) and Anganwadi workers are expected to be present in each village

Anganwadi covering a population of about a thousand. Anganwadi is a village-level

childcare centre set up under the Integrated Child Development Services program (ICDS). It

provides advice, information and basic health services for pregnant women, mothers and

young children. It has basic physical infrastructure and some training and skills in basic

health care. However, their presence in the village, their ability to build one-to-one

relationship with the women in villages and their ability to make home visits gives them an

advantage over both the formal health care delivery system and the traditional health care

practices that are still prevalent in India.


10/41

10

NCAER

State Health Department

FP

Clinics

RCH

ServicesImmunisations

Community

Health Centre

Rural

PrimaryHealth Centres

State Womanand Child Welfare

Department

Dispensaries Dispensaries

Male HealthWorker

Health SubCentre

FemaleANM

Households

Village LevelAnganwadiChild Care

Centres

Fig. 1: Structure of Delivery

of Health Care Services in India

HealthCare

Services

Family Planningand

Reproductive Health

Care Services

earch Medical

EducationHospitals


11/41

NCAERCommunity level facilities provided by the health department and the department of

woman and child welfare provide education and counseling on appropriate prenatal and

postnatal health care such as mother and child nutrition, need for rest and exercise, etc. They

also try to detect early danger signals and arrange for referral and help at the next level. The

primary and other higher level health care facilities provide immunisation services, detection

and referral for high risk factors and complications.

In the delivery of services, the community level help takes the form of conducting

aseptic deliveries with basic kits provided to the village level trained workers and providing

basic care like resuscitation for asphyxiated new born. In situations of a complicated delivery,

these workers refer the case to a higher order mostly from the Primary Health Centre or the

Community Health Centre in the rural areas. The workers at the health sub-centre supervise

home delivery, provide treatment for infections and deliver mother and child immunisations.

The paper analyses three aspects of maternity care services antenatal care, intra-natal or

child delivery and postnatal care that are described below.

Antenatal Care (ANC) and Postnatal Care (PNC)

Antenatal services are part of the primary health care services for pregnant women and

management of the foetus. In India ANC services consist of a set of professional pregnancycheckups, tetanus and other immunisations, prophylaxis through iron and folic acid tablets,

blood pressure check up and advice and information regarding delivery methods and services,


12/41

NCAERthat might have occurred during the delivery, severe health condition of the new born, etc. are

some of the PNC services that require skilled personnel and hospital facilities. The PNC

services provided at the community level include counseling on family planning, breast

feeding practices, nutrition, management of neo-natal hypothermia, early detection of

postpartum complications and referral for such problems. The higher-level health care

facilities are intended to provide these services as well as take care of post delivery

complications.

Health Care Services for Child Delivery

There are two aspects of the delivery services that are considered in this analysis whether

the delivery was at home or at a health care facility and whether a trained person was present

to assist in the delivery. Social norms in rural areas are such that home delivery is preferred to

institutional deliveries. This in itself is not a problem if hygienic and appropriate delivery

practices are used either by traditional helpers or by a professionally trained person who

makes home visits for helping with the delivery. Hence the government has set up a system in

which some village women, often the traditional midwives or dais, are given basic training

for child delivery. They are trained to deliver babies in a clean and proper way to avoid

infections and other complications. This has ensured that a large percentage of rural women

have access to trained help even for home deliveries.

Institutional Issues


13/41

NCAERIndia, the primary health and reproductive health care needs of the masses, especially in rural

areas, are expected to be provided by the public facilities, mostly run by the state

governments. However, due to various inefficiencies in the public health care delivery

system, even the minimum facilities are not often made available to the target groups in the

population.

In theory, the ANM should give pregnant women regular medical check-ups,

distribute iron and folic acid tablets (to combat anemia, which is presumed universal), give a

course of free injections of tetanus toxoid (to avert neonatal tetanus caused by cutting the

cord with an infected instrument), identify women at risk and assist in their deliveries or refer

them to the government hospital. The ANM should maintain a network of dais, usually those

she has helped to train, to identify pregnant women and register them with her. But in

practice most dais are not trained. Trained and untrained dais alike, seldom establish contact

with pregnant women and do not systematically deliver medical care to them. At the same

time, village women rarely consult dais during pregnancy. Most dais claim that they have no

treatments or medicines to administer so that women have no reason to bring their troubles to

them. Thus dais cannot provide a comprehensive safety net of medical care for pregnant

women.

The government antenatal services do not function well even in larger villages and

townsthat have an ANM clinic in the government dispensary. Tetanus toxoide injections are


14/41

NCAERcharges. Often the PNC services that require medication and treatment for lactating women

are denied to those who cannot pay.

The local primary health care centres and sub-centres are expected to provide most of

the ANC and PNC services free of cost and often at home. Although there is no explicit price

for most of these services, there is always an implicit price in the form of opportunity cost of

the time spent in availing health care facilities and services. Greater the distance to the

healthcare facility, higher the implicit price of the service. Thus physical accessibility

imposes another cost on the consumer, in addition to the illegal charges, for using a health

care system that is supposed to be almost free.

The demand related constraints originating from cultural practices and accessibility

factors, and the supply side constraints thus get reflected in the low overall rates of

reproductive health care utilisation in India. However, these averages hide considerable

variation in utilisation rates across different demographic groups and geographical regions of

the country. The next section examines the individual level data on reproductive healthcare

use to investigate this variation across groups and regions of the country.


15/41

NCAER3. DATA AND METHODOLOGY

The analysis in this paper uses data from the cross-sectional rural household survey

conducted by the National Council of Applied Economic Research in 1994 to prepare a

human development profile for India and its many states according to a number of

demographic groups (Sharrif, 1999). A two stage stratified sampling design was used to

sample about 32,000 families from 1765 villages across all parts of India. Household level

information such as income, assets, family demographics, etc. was collected along with

detailed information on the education and health of each family member. The health segment

of the dataset includes data on the reproductive health of adult women, the health of all

infants and details on the short and long-term morbidity of each family member. The

following analysis will use data on 7000 odd women in the reproductive age group who

delivered a child in the year before the survey.

The dimensions of reproductive health care considered in this paper are general ANC

use, blood pressure checkups, choice of home delivery versus institutional delivery, use of

professional help at the time of the delivery and finally the use of postnatal care. Although

the dataset has information on two other components of ANC iron tablet intake and

tetanus toxoid injections, we only use blood pressure checkups since this is a service that hasbeen introduced relatively recently in the ANC package.


16/41

NCAERwoman and preventive and curative health care use. These factors interact with the wages,

prices of different goods including health care and income from the budget constraint to

determine the demand for the reproductive health services. Several papers have presented a

detailed model of this approach.

As mentioned earlier although most of the ANC and PNC services are free of cost

there is an implicit price associated with utilisation of reproductive health services. This is the

opportunity cost of time spent on obtaining these services. It is equal to the wage rate for the

woman if she is in the labour force or the value of her time if she works at home. Thus, both

current own wages and current wages of other members (family income) should be

significant determinants of the demand for such services. Other services such as delivery at a

health care facility or trained help at the time of the delivery, although not explicitly priced,

have substantial direct expenditures. In these cases, household income and thus current wages

of all other family members become important determinants of health care demand. However,

we do not expect current own wage effect for the woman since, by the time of the delivery

the woman should be out of the labour force. Past wages of the woman could still affect the

demand of these services through a wealth (accumulated saving) effect in the current period.

On the supply side, we assume that the location decision for the family is notinfluenced by the availability of reproductive health care facilities. In practice, it is often

observed that a woman stays with her parents or parents-in-law, both for better home care


17/41

NCAER

The individual level variables considered in the analysis are the womans age and

education, the number of children and her work status. Since we do not have information on

past health status, education and wealth are used as proxies in the reduced form specification.

Ideally wages of all members of the household should be included as explanatory variables to

account for the implicit own and cross price effects associated with the opportunity cost of

time. However, due to data limitations explained later we only use average male and female

wages.

At the household level per capita income, land assets, number of adult females,

husbands education, caste and religion are included as explanatory variables. Womans

accessibility and exposure to information through media sources such as the radio, television

and newspapers and magazines is another household level variable. These are defined as

binary variables that take the value of one if any woman in the household frequently listens tothe radio, watches television or reads newspapers or magazines. The variable thus captures

the effect of the information exposure of women not just the mother. The supply side

variables and dummy variables for accessibility to hospital and health care service and the

indicators for the presence of an ICDS childcare in the village are listed in Table 1, which

provides detailed definition of the variables used in the analysis.

The analysis provides two specifications for the basic regressions. In the first


18/41

NCAERTable 2 presents some basic statistics from the data. Of the 37,649 currently married

women in the reproductive age group, 20 per cent delivered a child in the year before the

survey. The proportion of mothers who delivered a child has been higher in the states of

Rajasthan, Bihar and Uttar Pradesh. Of the total 7635 women, 61 per cent used some form of

ANC 33 per cent got regular blood pressure check ups, 10 per cent used iron prophylaxis

and 19 per cent got tetanus immunisations. Although we have information on iron intake and

immunisations, it is less reliable than the general measure of ANC use. Thus, the empirical

analysis only uses blood pressure check ups and ANC use as the two measures for antenatal

care.

A very large majority (75 per cent) of these deliveries occurred at home and only 52

per cent used some trained help at the time of delivery. The PNC use is only 26 per cent in

this sample. The women who used ANC were also much more likely to use other health care

services, for example, 37 per cent of the women who used ANC also used PNC. However,only 7 per cent of the non-ANC users used PNC.

These simple averages hide considerable variation across socio-economic groups and

regions with very different supply of reproductive health services that are discussed in an

earlier section. About 64 per cent of the women in this subgroup are illiterate. The mean per

capita income is Rs. 4068 for women in this group that is 9 per cent lower than the national

average of Rs. 4485. In the selected sample 29 per cent of the women reside in a village


19/41

NCAER4. EMPIRICAL ANALYSIS

The following section analyses the determinants of the five measures of maternity care

mentioned earlier blood pressure checkup, antenatal care, child delivery at home, use of

trained help at the time of the delivery and postnatal care. The direct effects of education,

information sources and health services are presented first, followed by the results from the

interactions among these variables. All probit regressions include age and age squared of the

woman, her primary occupation, state controls besides the variables included in the tables.

Basic Specification

Table 3 presents results from the first specification for the probit analysis of the utilisation of

maternity services. The determinants include demand and supply side factors such as

individual and household demographic characteristics, predicted per capita income and

access to health facilities. The instruments used to predict per capita income include land

ownership dummies, family size, ownership of farm assets such as farm machinery andtubewell, village level male and female wages. Table 4 presents similar results from a

specification where ownership of physical and human capital are included directly to

represent household income and wealth. Both tables present the coefficients only for the

important right hand side variables.

Education

These results show that womens education is a significant determinant of all measures of


20/41

NCAER

Husbands education also has a significant effect on most of the reproductive health

care utilisation measures. Matriculate education has the largest and statistically significant

impact on the probability of health care use. It increases the probability of pre and post natal

care use by about 10 per cent and 8 per cent respectively and the probability of the use of

trained help at the time of delivery by 7 per cent (Table 5). The magnitude of the effect of

husbands education is larger in the reduced form specification (Table 4), indicating that

husbands education affects utilisation directly through preferences and through increased

household income.

Household Composition

Next we examine the impact of some household composition variables on maternity care

utilisation. Most households in rural India are large multi-generation families with more than

one woman. The presence of other women in the household is important for a number ofreasons. They provide traditional maternity and childbirth information and advice from their

own experience as well as assistance at the time of delivery. For maternity health care

services home care can supplement or completely substitute outside medical care as in the

case of delivery related services. In addition, as the number of females in the family

increases, household income may also increase. Thus, the presence of women in the familyhas two opposing effects on health care demand the income effect that increases demand if

reproductive health is a normal good and a home care effect that families with more females


21/41

NCAERof children for the woman. Thus, this variable also captures the birth order effect that is

observed in other health outcomes for children. The large literature on childs welfare

health and education has indicated the significance of birth order in household decision

making in developing countries. Children born at lower parity have lower health and

education outcomes. In the case of maternity care utilisation we might expect that women

with greater number of children might use the past beneficial experience from health care use

to have greater utilisation rates. However the results are in the opposite direction. The birth

order effect could be due to a feeling of 'unwantedness' associated with births of higher order

or that the parents feel more experienced and less inclined to use professional help, especially

if there are explicit or implicit costs for this. However, we cannot completely separate the

two effects in this analysis.

Tables 3 and 4 show strong and significant birth order effects for the use of

reproductive health care. In the specification of Table 4, the variable captures both the birthorder effect as well as the effect of a lower per capita income. It shows that higher the

number of children, the lower is the probability of prenatal and postnatal care utilisation by

the mother and greater the chances of home delivery. Including per capita income allows for

the effect of children on household money resources but still does not control for the effect on

time and effort. The birth order or number of children effect exists even after controlling for

per capita income as shown in Table 3. Although significant, these effects are small in

magnitude with the effect of a unit increase in the variable increasing the probabilities by


22/41

NCAERwhich would be the current market wage for a woman in the labour force. Thus if the price of

the ANC or PNC is the womans current wage, a higher wage would imply less use of these

services. Wages of other family members should be included to allow for substitution of

labour supply between the pregnant woman and other family members.

Unfortunately the dataset does not provide individual level labour supply or current

wage information. Instead, we use gender specific village level wages, averaged over

agricultural and non-agricultural activities, as the predicted current wage for men and women

in the sample. In the probit regression, the coefficient of female wages in the regression

represents the opportunity cost of time in using these free services. The average male wage

represents the cross-price effects and captures substitution of female labour supply by male

labour supply. Thus, an increase in male wages has a positive effect on utilisation rates, via a

substitution of female work by male work and a pure income effect.

On the other hand, hospital delivery and the use of trained help at the time of delivery

have explicit costs. Thus, male wages would affect the demand for these services via

household income while, conditional on the fertility decision, we do not expect current

female wage to affect the decision on place of delivery or the use of trained help. There is no

avoidable opportunity cost of time associated with using these services. The reduced form

specification in Table 4 uses physical and human capital and agricultural wages as proxies for

household income. The average male wage always has a significant effect and increases the


23/41

NCAERincome effects for the delivery services since irrigation substantially increases agricultural

output and income.

Table 3 shows the direct effect of per capita income, instrumented using physical and

human capital and wages. The results show that income is a significant determinant of the

probability of home delivery and the probability of using trained help at the time of the

delivery but does not affect the use of antenatal and postnatal care. However, the coefficients

on land ownership indicate that, after controlling for per capita income, women from the land

and business owning households are less likely to use ANC than those from the landless

agricultural labour class. In India cultural and traditional practices often isolate pregnant

women and restrict their physical movements. These are more likely to be observed in the

landed and upper income categories in the rural areas and could lead to these results. The

lower restrictions on womens movements in the wage earning households and their greater

participation in the labour force makes it easier for them to seek professional health careservices relative to other women.

An argument can be made that the average wages are capturing the development level

of the villages rather than the opportunity cost of time. However, that does not explain the

opposite signs on the coefficients of male and female wages. It is true that since we do not

have actual wages and labour force participation these results are only indicative of wage

effects but they point to the need for further research in this area. These results are also


24/41

NCAERno woman did. The effects of newspaper reading are minimal and significant only for the

choice of place and method of delivery.

Transport Facilities

As mentioned in the context of wage effects, it seems that the opportunity cost of time is an

important issue in determining health care use. This could be due to the physical distance to

the health care centres so that means of transport, both public and private, become important

in influencing decisions to seek prenatal and postnatal health care. The results show that

ownership of vehicles does not have any effect on the utilization of health care but access to

public transport is a significant determinant of the use of ANC and PNC. Greater the distance

to the closest bus service lower the use of prenatal and postnatal care. Since public transport

is unlikely to be used at the time of delivery, it is excluded from the regressions for child

delivery services. However, this variable could also be picking up general development level

of the village that reflects general awareness, exposure and facilities in the village.

Health Care Services

The coefficients on the health care facilities show that the most effective institution for

antenatal care has been the ICDS (Anganwadi). Their local presence in the village and the

personalised interactions of the staff has made them an important source of basic

reproductive health care. This has important policy implications in terms of expansion of

health facilities that are most effective. The presence of an Anganwadi in a village increases


25/41

NCAERhealth workers. These results indicate a need for more focused programs to bring these

communities at par with others. Similarly, large state level differences are also observed in

the use of these health services that reflect state specific factors. These could be the demand

side factors in terms of general education, information and awareness as well as the supply of

such services, both in quantity and quality.

Education, Information and Health Interactions

The results presented above point to the significant impact of education, information and

health care access on health care utilisation. In this section we examine some of the

mechanisms through which these effects operate, e.g., does the education effect work through

better information processing. We look at some of the interactions between the main

explanatory variables to answer questions such as whether female education and media

information are complements or substitutes or does higher education result in more health

care usage for areas with better health facilities?

In order to examine how accessibility to health care centres affects health care use as

market value of time increases, Table 7 presents results from a model with interaction

between womens wages and access to health care facilities. Since this wage variable was

not significant in the utilisation of child delivery service, the results presented here are only

for prenatal and postnatal care services. In all cases the effect of female wage declines

substantially in magnitude and becomes statistically insignificant. The direct effect of health


26/41

NCAERNext, we consider the interactions between the womans education and the

information processing variables to see if one enhances the effect of the other. The results in

Table 8 show that the introduction of the interaction terms does not affect the significance of

the education terms in most cases. The magnitude of the direct effects are mostly unchanged

for primary education but decline for middle and even more for matriculate education for

almost all cases of maternity care use. The direct effects of the information variables are

reduced in all cases except for the utilisation of postnatal care. The introduction of

information and education interactions reduces the magnitude of the pure effect of these

variables for ANC use and increases the pure effects for PNC usage.

Middle and higher education increase the effectiveness of the media in increasing the

order to examine regional variations in maternity care utilisation, we repeated the above basic

analysis for the utilisation of antenatal care and skilled help at the time of the delivery while

decreasing the incidence of child delivery at home. Thus, strong complementarity existsbetween female middle and higher education and exposure to media. This in turn implies that

the more educated women are better able to process and assimilate the information provided

through the radio, television or the printed media. However, there are no significant cross

effects in the utilisation of postnatal services. The interaction terms are jointly significant at

the 5 per cent level only for the utilisation of antenatal care and home delivery.

Finally, Table 9 presents results from regressions that include interactions between


27/41

NCAERlabour force status of the woman. However this could be because the dummy does not

adequately represent current labour force participation status for the pregnant woman.

An attempt is also made to study the regional differences. Broadly, it may be stated

that general utilisation rates are higher in the south. The magnitude of the effects of key

determinants such as education and health care access variables is significantly different for

these regions. The educational effects for lower levels of education are stronger in the north

reflecting a greater marginal gain to be made by improving education levels in the north. This

could be due to lower average level of education in the north relative to the south. Education

effects are stronger for delivery variables in the south and PNC in the north.


28/41

NCAER5. CONCLUSIONS

One of the surest ways to improve the quality of life and human development amongst the

masses in India is to provide for quality primary health care services especially in the rural

areas. Since the health care services are concentrated in urban areas, the rural area programs

should improve the outreach activities relating to health care delivery. This scheme is all the

more essential in providing the reproductive care services that are related to mother and child

care such as the ANC, natal and PNC services. Although the demand for such services are

affected by a number of cultural stereotypes that are still prevalent in society, improving

quality of services, making them easily available, providing them in their own villages and

free of cost is absolutely essential. Since often pregnancy is not regarded as a condition that

warrants medical and clinical attention new strategies have to be evolved to ensure that

millions of pregnant and lactating mothers receive the benefits of modern technology that

will reduce both maternal and infant and child mortality that is essential to speeding up thedemographic transition towards low fertility and population stabilisation.

More specifically the use of reproductive health care such as antenatal and postnatal

care and child delivery with the help of trained personnel greatly improves the chances of

survival for the newborn. It results in healthier and stronger infants leading to good health in

later life. It also leads to better health for the mother that in turn implies better health for the

child. However, empirical research on the determinants of reproductive health care use has


29/41

NCAERAccess to locally available health services significantly increases maternity care use. An

important health care facility in this respect has been the village level ICDS center.

The results also show that information sources such as radio and television are

complementary to womens education in increasing the utilisation of health care. However,

the interactions between education and access to health care facilities are insignificant except

for the complementary relationship between higher education and access to health care

facilities for PNC use.

These results point to the avenues through which policy makers can affect the

utilisation of health care. That greater education for women and greater use of media to

spread information, which work through changing preferences, enhance demand and

utilisation is the prominent finding of this study. The supply side dimension such as the

imminent expansion of local health care facilities, deserves the urgent attention of the policymakers. Policy efforts in improving education, spreading health care information through the

media and providing a better local network of health workers will go a long way in

augmenting development through improving the health of the mother and the child.


30/41

NCAER

REFERENCES

Jeffery, Patricia, Roger Jeffery and Andrew Lyon (1988), Labour Pains and Labour Power:Women and childbearing in India, London: Zed Books Ltd.

Measham, Anthony R. and Meera Chaterjee (1999), Wasting Away: The crisis of malnutritionin India, Washington DC: World Bank.

Nayler, David C., Prabhat Jha, John Woods and Abusaleh Shariff (1999), A Fine Balance:Some Options for Private and Public Health Care in Urban India,Washington DC: World

Bank, pp i ix and 138.

Shariff, Abusaleh (1990), A Few Cultural Concepts and Socio-behavioural Aspects of

Human Health in India, in What We Know About Health Transition: The Cultural, Social andBehavioral Determinants of Health (eds. J.Caldwell et al.), pp. 788805,The Proceedings of

an International Workshop, Canberra: The Australian National University.

Shariff, Abusaleh (1993), Determinants of Child Health: Search for Maternal Education

Effects in Gujarat, Working Paper No. 47, Ahmedabad: Gujarat Institute of DevelopmentResearch.

Shariff, Abusaleh (1999), INDIA: Human Development Report, New Delhi: OxfordUniversity Press, June, pp. ixiii and 1370.


31/41

NCAER

TABLE 1Variables and Definitions

Community Level Variables

Bus Far Dummy Variable=1 if bus service beyond 9 kms of the village

Bus Medium Dummy Variable=1 if bus service within 9 kms of the village

Bus Close Dummy variable=1 if bus service in the village

Health worst Dummy variable=1 if health centre not in the village and hospital > 5 kms from thevillage

Health Medium Dummy variable=1 if health centre not in the village and hospital < 5 kms from thevillage

Health Best Dummy variable=1 if health centre in the village and hospital < 5 kms from thevillage

Anganwadi Dummy Variable=1 if the village has an ICDS centre ( Integrated ChildDevelopment Services)

Household Level Variables

Land 0 Dummy variable=1 if landless worker

Land 1 Dummy variable=1 if small or medium size farmer

Land 2 Dummy variable= 1 if large farmer

Land 3 Dummy variable =1 if landless and own business

Vehicle 0 Dummy variable =1 if no vehicle owned

Vehicle 1 Dummy variable =1 if family owns a cycle or bullock cart

Vehicle 2 Dummy variable =1 if the family owns a tractor or car


32/41

NCAERMiddle Dummy variable=1 if 8 years of schooling

Matriculation Dummy variable=1 if 12 or greater years of schooling/college

TABLE 2

Sample Means

VariablesAll Women

( 7635)

Users ofReproductive


( 5816)

Non-Users ofReproductive


(1819)

TT Immunisation 19 %

Iron Intake 10 %

Blood Pressure Checkup 33 %

ANC use 61 %

Delivery at Home 75 %

Trained Help for Delivery 52 %

PNC use 26 %

Illiterate 64 % 57 % 83 %

Primary Education 18 % 21 % 10 %

Middle Education 10 % 11 % 5 %Matriculate and Higher

Education 8 % 10 % 2 %

Poor Health Facility 31 % 28 % 39 %

Medium Health Facility 40 % 40 % 41 %


33/41

NCAERTABLE 3

Reproductive Health Care Utilisation1

Variables Blood PressureCheck Up

Antenatal Care Home Delivery Delivery withTrained Help

Postnatal Care

Education (Self)Primary 0.299* 0.265* -0.412* 0.181* 0.292*

Middle 0.317* 0.116*** -0.486* 0.339* 0.232*

Matriculation 0.631* 0.336* -0.573* 0.243* 0.258*Education (Husband)Primary 0.126* 0.105* -0.001 0.087** 0.051

Middle 0.087 0.126* -0.084 0.133* 0.079

Matriculation 0.161* 0.294* -0.160* 0.153* 0.233*

Number of females 0.011 -0.014 0.032*** -0.034** -0.026

Number of children -0.084* -0.048* 0.124* -0.024 -0.054*

Land 1 0.075 -0.119* 0.087 -0.032 -0.060

Land 2 0.015 -0.071 -0.115** 0.122* 0.007

Land 3 0.137** -0.095*** -0.141** 0.005 -0.097Predicted Per CapitaIncome ( Log) - 0.149 0.080 -0.307* 0.280* 0.018Information VariablesRadio 0.152* 0.131* -0.074*** -0.008 0.107*

Television 0.155* 0.105* -0.109* 0.179* 0.074***

Newspaper/Magazines 0.004 -0.005 -0.142* 0.074 0.086Transportation

Vehicle 1 0.090** 0.043 0.026 -0.034 0.013

Vehicle 2 0.205* -0.032 -0.004 -0.067 0.040

B S i (M di ) 0 191* 0 180* 0 208*


34/41

NCAERTABLE 4

Reproductive Health Care Utilisation2

Variables Blood PressureCheckup


Postnatal Care


Middle 0.312* 0.111*** -0.471* 0.331* 0.231*

Matriculation 0.637* 0.345* -0.550* 0.223* 0.260*Education (Husband)Primary 0.126* 0.109* -0.008 0.095* 0.058

Middle 0.083 0.137* -0.103*** 0.157* 0.089

Matriculation 0.155* 0.314* -0.200* 0.204* 0.259*

Number of females 0.030 -0.005 0.008 0.010* 0.001

Number of children -0.073* -0.055* 0.150* -0.047* -0.056*

Income and Wealth ProxiesLand 1 0.039 -0.087** -0.007 0.055 -0.038

Land 2 0.019 -0.064 -0.122 0.130* 0.023

Land3 0.123* -0.081 -0.161 0.025 -0.100

Farm machine 0.013 0.051 0.089 -0.124 -0.137

Tubewell -0.085 0.006 -0.201* 0.171* -0.113***

Average Male Wage 0.006** 0.008* -0.010* 0.008* 0.008*

Average Female Wage -0.011* -0.012* 0.004 -0.002 -0.005*Information VariablesRadio 0.142* 0.123* -0.066 -0.010 0.103*

Television /Cinema 0.152* 0.105* -0.100* 0.180* 0.078***

Newspaper/Magazines 0.0003 -0.009 -0.137* 0.075 0.081TransportationVehicle 1 0.063 0.055 -0.033 0.030 0.031


35/41

NCAERTABLE 5

Reproductive Health Care Utilisation Changes in Predicted Probability3(%)

Variables Blood PressureCheck Up


Postnatal Care


Middle 11.6* 4.3*** -15.9* 13.0* 7.6*


Middle 3.0 4.6* -2.4 5.2* 2.5

Matriculation 5.7* 10.6* -4.7* 6.0* 7.5*

Number of females 0.3 -0.5 0.9*** -1.4* -0.8

Number of children -2.9* -1.8* 3.5* -0.9 -1.7*

Land 1 2.6 -4.5* 2.4 -1.3 -1.8

Land 2 0.5 -2.7 -3.3** 4.8* 0.2

Land 3 4.8** -3.6*** -4.2** -0.2 -2.9Predicted Per CapitaIncome ( Log) -5.1 -3.0 -8.8* 11.1* -0.6Information VariablesRadio 5.3* 4.9* -2.1*** -0.3 3.3*

Television 5.4* 3.9* -3.2* 7.1* 2.3***

Newspaper/Magazines 0.1 -0.2 -4.2* 2.9 2.7Transportation

Vehicle 1 3.1** 1.6 0.7 1.3 0.4

Vehicle 2 7.4* 1.2 -0.1 2.7 1.2

B S i (M di ) 6 4* 6 8* 6 1*


36/41

NCAERTABLE 6

Reproductive Health Care Utilisation Changes in Predicted Probabilities4(%)

Variables Blood PressureCheckup


Postnatal Care


Middle 11.4* 4.1*** -15.3* 12.7* 7.5*


Middle 2.9 5.0* -3.0*** 6.2* 2.8

Matriculation 5.5* 12.1* -5.9* 8.0* 8.3*

Number of females 1.0 -0.2 0.2 0.4 0.04

Number of children -2.5* -2.1* 4.2* -1.9* -1.7*Income and Wealth ProxiesLand 1 1.3 -3.3** -0.2 2.2 -1.2

Land 2 0.7 -2.4 -3.6 5.1* 0.7

Land3 4.3* -3.1 -4.8 1.0 -3.0

Farm machine 0.5 1.9 2.4 -5.0 -4.0

Tubewell -2.9 0.2 -6.1* 6.7* -3.4***

Average Male Wage -0.2* 0.3* -0.3* 0.3* 0.2*

Average Female Wage -0.4* -0.4* 0.1 -0.1 -0.2*Information VariablesRadio 4.9* 4.6* -1.9 -0.4 3.2*

Television /Cinema 5.3* 3.9* -2.9* 7.1* 2.4***

Newspaper/Magazines 0.01 -0.3 -4.0* 3.0 2.5TransportationVehicle 1 2.2 2.1 -0.9 1.2 0.9

Vehicle 2 4 5*** 0 3 6 0* 6 0* 4 4


37/41

NCAERTABLE 7

Reproductive Health Care Utilisation Wage and Health Services5

Blood Pressure Antenatal Care Postnatal CareVariables

WithoutInteractions With Interactions WithoutInteractions With Interactions WithoutInteractions With Interactions

Income and Wealth ProxiesLand 1 0.039 0.040 -0.087** -0.086*** -0.038 -0.032

Land 2 0.019 0.019 -0.065 -0.065 0.024 0.020

Land3 0.123* 0.125* -0.081 -0.078 -0.100 -0.095

Farm machine 0.013 0.011 0.051 0.050 -0.137 -0.147

Tubewell -0.085 -0.089 0.006 0.003 -0.113*** -0.112***

Average Male Wage 0.006* 0.006* 0.008* 0.008* 0.008* 0.007*

Average Female Wage -0.011* -0.007*** -0.012* -0.004 -0.005* -0.002Health ServicesMedium

Best

-0.051

0.129*

-0.044

0.173

-0.015

0.146*

0.102

0.286*

0.010

0.189*

-0.038

0.133

Anganwadi 0.060 0.223* 0.132* 0.249* 0.037 0.210*Health Medium x FemaleWage 0.0001 -0.004 0.002

Health Best x FemaleWage -0.0015 -0.005 0.003

Anganwadi x Female Wage -0.0063** -0.004 -0.007*

P >X2

(interactions) 0.268 0.238 0.072


38/41

38

NCAERTABLE 8

Reproductive Health Care Utilisation Education and Information Processing

Blood Pressure Antenatal Care Home Delivery Delivery with Trained Help Postnatal CareVariables

Without

Interactions

With Interactions Without

Interactions


Interactions


Interactions


Interactions

With Interactions

Land 1 0.075 0.077 -0.119* -0.123* 0.087 0.086 -0.032 -0.037 -0.060 -0.057

Land 2 0.015 0.018 -0.071 -0.072 -0.115** -0.120 0.122* 0.125* 0.007 0.008

Land3 0.137** 0.137** -0.095*** -0.099*** -0.141** -0.140** 0.005 0.004 -0.097 -0.094

Predicted Per Capita Income(Log) - 0.149 -0.144 0.080 0.086* -0.307* -0.320* 0.280* 0.289* 0.018 0.025

Education ( Self)

Primary 0.299* 0.282* 0.265* 0.240* -0.412* 0.430* 0.181* 0.196* 0.292* 0.287*

Middle 0.317* 0.192*** 0.116*** 0.129 -0.486* 0.300* 0.339* 0.278* 0.232* 0.356*

Matriculation 0.631* 0.391* 0.336* 0.170 -0.573* 0.401* 0.243* 0.134 0.258* 0.277**Education (Husband)

Primary 0.126* 0.134 0.105* 0.106* -0.001 0.005 0.087** 0.086** 0.051 0.046

Middle 0.087 0.101*** 0.126* 0.127* -0.084 0.098 0.133* 0.137* 0.079 0.075

Matriculation 0.161* 0.167* 0.294* 0.295* -0.160* 0.164* 0.153* 0.148* 0.233* 0.227*

Information Processing

Radio 0.152* 0.109* 0.131* 0.114* -0.074*** 0.094 -0.008 0.027 0.107* 0.124*

Television 0.155* 0.131* 0.105* 0.069 -0.109* 0.034 0.179* 0.109* 0.074*** 0.091***

Newspaper/Magazines 0.004 -0.028 -0.005 0.102 -0.142* 0.057 0.074 0.077 0.086 0.053

Radio X Primary -0.020 -0.046 0.144 -0.134 -0.022

Radio X Middle 0.280* -0.038 -0.152 -0.056 -0.089

Radio X Matriculation 0.234 0.479* 0.006 -0.020 -0.030

Television X Primary 0.104 0.147 -0.140 0.210* 0.040

Television X Middle -0.046 0.157 -0.278** 0.202 -0.143

Television X Matriculation 0.066 -0.173 -0.051 0.027 -0.020

Newspaper X Primary -0.066 -0.124 0.056 -0.200 -0.036

Newspaper X Middle -0.013 -0.309*** 0.182 -0.021 0.126

Newspaper X Matriculation 0.161 -0.117 -0.401** 0.262 0.088

P > X2(interactions) 0.263 0.055 0.021 0.146 0.952


39/41

39

NCAER


40/41

40

NCAERTABLE 9

Reproductive Health Care Utilisation Education and Health Services

Blood Pressure Antenatal Care Home Delivery Delivery with Trained Help Postnatal CareVariables

Without

Interactions


Interactions


Interactions


Interactions


Interactions

With Interactions

Land 1 0.075 0.021 -0.119* -0.121* 0.087 0.088 -0.032 -0.016 -0.060 -0.063

Land 2 0.015 0.008 -0.071 -0.076*** -0.115** -0.117** 0.122* 0.123* 0.007 0.007

Land3 0.137** 0.117** -0.095*** -0.097*** -0.141** -0.137** 0.005 0.013 -0.097 -0.090

Predicted Per Capita Income(Log) - 0.149 0.029 0.080 0.089 -0.307* -0.297* 0.280* 0.222* 0.018 0.051

Education (Self)

Primary 0.299* 0.395* 0.265* 0.324* -0.412* -0.389* 0.181* 0.177*** 0.292* 0.330

Middle 0.317* 0.227 0.116*** 0.136 -0.486* -0.462* 0.339* 0.169 0.232* 0.051

Matriculation 0.631* 0.550 0.336* 0.368** -0.573* -0.300** 0.243* 0.449* 0.258* 0.521Education (Husband)

Primary 0.126* 0.125 0.105* 0.108* -0.001 0.001 0.087** 0.086** 0.051 0.047

Middle 0.087 0.074 0.126* 0.125* -0.084 -0.082 0.133* 0.140* 0.079 0.077

Matriculation 0.161* 0.147 0.294* 0.298* -0.160* -0.162* 0.153* 0.159 0.233* 0.228

Health Services

Medium -0.038 -0.078 -0.004 -0.179* -0.035 -0.040 -0.009 0.003 0.016 -0.039

Best 0.135* 0.147* 0.143* -0.237* -0.136** -0.064 0.160* 0.156 0.193* 0.177*

Anganwadi 0.068*** 0.094** 0.143* 0.005 -0.021 -0.009 0.039 0.029 0.050 0.092***

Health Medium X Primary -0.027 -0.054 0.055 -0.091 0.045

Health Medium X Middle 0.166 -0.056 0.015 0.243 0.316**

Health Medium X

Matriculation

0.342 0.006 -0.176 -0.231 0.012

Health Best X Primary -0.004 0.029 -0.178 -0.038 0.080

Health Best X Middle 0.027* -0.184 -0.114 0.187 0.368*

Health Best X Matriculation -0.007 -0.086 -0.217 -0.111 -0.240

Anganwadi X Primary -0.143 -0.085 0.029 0.095 -0.148

Anganwadi X Middle 0.062 0.121 0.019 0.015 -0.126

Anganwadi X Matriculation -0.098 -0.031 -0.214*** -0.118 -0.233**

P > X2 (interactions) 0.202 0.879 0.423 0.563 0.043


41/41

41

NCAER

determinant india

Documents